Dual Disk Clutches

[The Dark Side of Will]

Playing with flywheels and taking some measurements to design the flexplate.

PTT on the left, QMI on the right. QMI should be much easier for adding the flexplate.

Oh yeah, good thing I measured. The QMI flywheel is actually 0.837 or so... a little short even for a nominal 0.850, much less 0.890.

[The Dark Side of Will]

Machinist made a first article Throw Out Bearing Holder from one of my blanks, but it needed some tuning up. I got that done today.
I think I have the throw out bearing holder's interface on the transmission side finalized. Now I just need to make one last trip through the stack-up for the clutch to make sure I have the right height for holding the bearing, then get that made.

[The Dark Side of Will]

Was able to get out to see the machinist today. He's had a busy summer so far; took the family on a vacay to Montana a couple weeks back.

I had a fancy way figured out to modify any old 142 tooth flexplate (well... any flat enough 142T flexplate) to bolt to the back of the PTT flywheel. That's a complicated fussy bit of machining and the unit/assembly would need to be balanced afterward.
I sharpened the pencil a bit and determined that even with the worst case situation of the crank flange being 0.020" proud of the bellhousing face, I could stack the PTT flywheel on top of a 0.100" thick flexplate and still have 0.035 clearance to the Getrag output bearing boss inside the bellhousing. I measured my engine and the crank flange is actually 0.012-0.015 BELOW the bellhousing surface, so stacking the flywheel on top of the flexplate should leave me with 0.065 or so between the clutch and the transmission case.

To that end, I checked my 11mm Cadillac flexplate against the PTT and TIlton flywheels and the BOTH FIT PERFECTLY.



So the only thing that has to be done is drill the crank bolt circle and bore the crank pilot bore into the flywheel. I left him my spare forged crank to use for fitment. The crank pilot is only a smidge thicker than the flexplate, so an alignment tool may be necessary to locate the flywheel on the shallow hole drilled into the center of the crank pilot.

Hopefully that'll be done next weekend and I can get the clutch bolted up. Then it'll be on to the throw out bearing holder and THEN mating the engine and trans!
I'm definitely using modeling clay to make sure my clearance estimate is good, though.

[The Dark Side of Will]

Flywheel was ready this weekend, so I got a BUNCH of test fitting done.

Stock forged crank flexplate for use with PTT flywheel





I told the machinist to leave a sharp corner on the back side of the crank pilot bore, since the Northstar crank pilot is very short. He did that and stoned the back surface to deburr... Also leaving sharp corners on the bolt holes. I hit those with a countersink, while keeping the sharp corner on the pilot bore.



Lol... I made spacers and snagged longer bolts for the engine stand. While I may be able to install the engine to the stand with clutch, it turns out I still need to remove the engine from the stand in order to install the clutch.





ACT ATGM14 clutch alignment tool worked like a charm, although it's a tight fit into the Northstar crank and difficult to install.
On Summit: https://www.summitracing.com/parts/acl-atgm14



A key problem I noticed early is that some of the bolt heads hit the flexplate. More on that in a minute.



Here's the transmission fitted:



This is with the stock TOB holder installed inside the bellhousing. This dimension should come down about 0.350" once I get my new TOB holder finished.



Took all the bolts out and it didn't pop off on its own, so nothing is binding.



I've been forgetting to do the one slight block mod required to bolt a Northstar up to a conventional Metric Bellhousing transmission... so here it is graphically:



Here's how flat the diaphragm spring fingers are when the unit is bolted all the way down. This is why I need a taller than stock TOB holder.



The throw out travel is 0.300", and all the manufacturers caution strongly against overtraveling throw out, as that can damage the clutch. I guess the damage is not from the diaphragm spring fingers contacting the disks (possibly from stretching the spring too far?)



I applied strip caulk to the transmission bellhousing. After one failed attempt, I got a measurement by applying the wax paper packing on top of the caulk so it wouldn't stick to the pressure plate cover. You can see the speed holes in the cover in relief in the caulk. This sample calipered at 0.080, so I have even a bit more clearance than estimated. Yay for conservative design and measurements.



Above I mentioned that the bolt heads hit the flexplate. Once I completed the clutch test fit, I looked into that. The anti-rotation features on the bolt flanges all cleared the flexplate. However, this one was REALLY close. That's a 0.005 feeler gauge that didn't make it quite all the way through. A 0.004 feeler will go all the way through, but held by friction.

In this shot, I had already skimmed the bolt heads to clear.



Now, check this out:

This is 180 opposite the one above



Gauge pins tell me that gap is between 0.050 and 0.054. That's almost FIFTY thou larger than the gap on the other side. Considering the tolerances of the flywheel, I think the difference is ENTIRELY in the flexplate. That's 0.050 of axial TIR across the stamped metal part. Keep in mind that GM bolted a 4 lug torque converter that was considerably flatter than that up to this flexplate and... sent it. Enjoy your Cadillac, Sir.

I'm not saying GM's tolerances are crap... This part is a potato chip, but it's a FLEXIBLE potato chip compared to a torque converter. Just pull it flat with the bolts and apparently it lasts well enough that the 4L80E transmissions acquired a reputation for being consummate anvils. Absolute units. Bulletproof. Better is the enemy of good enough, and GM knows EXACTLY where good enough is for this critical interface. Another aspect of this is that the ring gear was probably flat when it had a potato chipped stamping welded to it. That means when the potato chipped flexplate is pulled flat on the converter, the ring gear now has axial runout... and apparently that does not matter either.

Here's fit checking the first article TOB holder on my actual transmission



Here's a video of cycling the same. I didn't realize it until after I sat down to write this post, but it reminds me of something. I'm trying to remember what.

https://i.imgur.com/oFZpDHd.mp4

Bonus shot of the flywheel. Yeah, I was just using 4 bolts for the fit checks. I'll use 8 for the flight assembly. Those are the 0.880 long LS flexplate bolts. Those bolts are the PERFECT length when using only the flywheel. When stacking the flywheel on top of the flexplate, now they have 0.500 thread engagement in a crank flange that's 0.570 thick. I need to call ARP to determine if that's ok or if I have to buy the 1.075 long bolts and trim them down to 0.980 or so. A 25mm bolt would be the PERFECT length for this configuration, but ARP doesn't make any 25mm 11x1.5 flywheel bolts.

[The Dark Side of Will]

Somehow I missed this glamor shot on my prior upload:

[The Dark Side of Will]

WOOOOOooOO00OOOooT!

ARP wants the flywheel bolts lubricated with assembly lube under the heads, but also to have blue Loctite on the threads. Of course you can't get assembly lube on the threads, or that will screw up the Loctite, so you have to be a bit... careful.
Laid out the flywheel & flexplate for goo application per ARP's instructions... assembly lube under the heads in this photo



Now with Loctite applied



From there I just picked up the block the "assembly" was sitting on and pushed it up against the engine while holding the parts so they don't fall and using my 3rd and 4th hands to get the bolts started.
But here they are, "installed for flight"



I let the Loctite cure overnight, then sprayed the face down with brake cleaner and then isopropyl alcohol, trying to wash the assembly lube out of the bolt head interface, so that centrifugal force doesn't flow it out into my clutch friction interfaces. That of course damaged the paint, so I won't post pics of that.

And here's the flight install of the clutch, although not nearly as permanent as the flight install of the flywheel & flexplate.
The aforementioned installation tool worked great. I put a little more time into cleaning out the pilot bore in the crank, which still had a little scunge from the balance job inside it.
Damn, I finally have a clutch!



Shot to verify which direction I installed the disks that turned out unexpectedly artsy



The tips of the bolts are a smidge proud of the top of the clutch, but based on my measurements last week, this should be ok.



I got the engine back on the stand and used a 1/2" round file to perform the singular block mod required by this swap:



And here's another shot on the engine stand before I put it away for the week. Yes, I bag it to keep it clean.



Here's a little dorking around with the clutch alignment tool.
Still reminds me of something, but I can't think of what...
Not sure if these come through with sound. Hmmm...

https://i.imgur.com/mSV1DTl.mp4

https://i.imgur.com/9jE7KDN.mp4

[The Dark Side of Will]

Now that I have established that the flywheel+flexplate (fly+flex? flyflex? Sounds like I should (TM) that...) stack just barely fits into the 282 bellhousing.

Here's a pic I grabbed from the Pioneer catalog (FRA-571):



This is the flexplate for the cast crank Northstars. It has 8mm crank bolts. As near as I can tell from basic inspection, the only difference between this and the flexplate for the forged crank is the size of the crank bolt holes.

The contour of the stamped part of this flexplate is what allows the flywheel to be stacked on top of it. Notice in the center the surface for the crank flange is fairly large; the bolt circle is 78mm (3.070"). Moving outward, there's a narrow flat ring around the surface for the crank flange. That ring is actually closer to the engine than the surface for the crank flange is. Outward from that flat ring is a very shallow cone opening away from the engine. That cone blends into the outermost flat (well... not *that* flat) ring which carries the 4-bolt pattern for the torque converter. Notice the upper right hole is slotted.

There are two potential points of interference vs. a button flywheel for a 7.25" dual disk clutch.

Here's a photo from above of a flexplate sitting on a flywheel:
The first potential interference point is just around the narrow flat ring where the crank flange of the flywheel is still flat.



I did NOT measure that dimension on the PTT flywheel before I bolted it on, so I need to email PTT about that to see what's up.
The PTT looks like a good choice in this case, as the crank flange surface is *slightly* smaller than the crank flange surface on the Tilton flywheel.



The second potential interference point is at the outer flat ring vs. the bolt lugs at the perimeter of the flywheel. The shallow cone has to be shallow enough to miss the bolt heads, as depicted here:



This is the interesting part... From measuring V6 and Northstar flexplates to figure out where my ring gear needs to be before I realized I could use an intact flexplate, there are slight differences in the relationship of the ring gear to crank flange between the Caddy and V6 parts. I guess I also need to add LS4 flexplate data to that pile.

The LS4 flexplate can tolerate a flywheel crank flange contact surface up to 4.125" diameter. While I didn't measure (or maybe just didn't write down) the contact diameter on the PTT flywheel, I know it's larger than 4.125 because I tried to fit the PTT flywheel to the LS4 flexplate and it did NOT fit. That combo would require a spacer between the flexplate and the flywheel, which means it would not work in a 282 bellhousing... but nobody's trying to use a 282 with an LS4 anyway.

[Shaun41178(2)]

I planned to use my 282 with the ls4 I had here. I wasnt planning on going twin disk though. Was only gonna make about 400 hp or so so twin wasnt needed imo

[The Dark Side of Will]

Well we already know you're a weirdo.

I will believe the PTT clutches are as streetable as they claim once I drive one. Until then I'm skeptical.

My SPEC Stage 3 was a chattering nightmare in traffic. My Centerforce before that was simply inadequate for a V8. I haven't tried a SPEC Stage 3+, and while have heard they're better than the nonplussed Stage 3, the full-faced disk slows down shifting, especially with the 282's single cone synchros. I heard good things about ClutchNet when they first got started, but I heard bad things about ClutchNet once they'd been in business for a few years.

I'm just fed up with the offerings in the stock type 9 1/8" to 9 11/16" architecture & form factor. They seem to range from unacceptable to compromised while occasionally being built by nincompoops who assemble disk hubs backwards. There's never been a product in that range that makes a powerful Fiero's clutch not a problem.

[The Dark Side of Will]

The machinist got the TOB Holder back to me... I didn't play with it much because this turned into a prep & paint weekend. I need to noodle the design a little more to figure out if I have it right or not... I probably should order a back up bearing in case I need to drive this one back off after the test fit. The circle of 4 holes is for driving the bearing off. This unit places the bottom of the TOB 0.350 closer to the engine than the stock stamped holder/carrier would.

[ericjon262]

do you have any concerns with the flex plate flexing and contacting the lugs on the flywheel? could trigger the knock sensors.

[The Dark Side of Will]

We'll see? It can't not have some a circumferential harmonic mode, but I don't know if the displacement would be as much as 0.005" in the middle of the radius like that... especially since there's no axial loading to induce a deflection and the only operating force would be said circumferential mode.

Bets I can squeeze some foam rubber in there with tweezers and an inspection mirror without pulling the trans off the engine?

[pmbrunelle]

I was shopping for anti-vibration foam rubber.

Rogers Poron 4701 was recommended to me by a vendor. I liked it; it has a good squishiness for anti-vibration. I am trying it in some prototypes at work.

The squishiness of some closed-cell foams depends on the gas bubbles trapped inside. As the gas slowly diffuses out, the foam loses its firmness.

I don't fully remember the explanation, but the Poron's stiffness doesn't depend on trapped gas bubbles; it comes from the polyurethane itself, so it is immune to the gas diffusion issue.

[The Dark Side of Will]

I assembled the TOB & Holder. The press fit of the TOB to the hold squeezed down the bushing to the point that it's no longer a slip fit on the TOB guide in the transmission. Interesting... I'll have to figure out what to do about that.

[Shaun41178(2)]

Time for a 6 month update

[ericjon262]

Time for a 6 month update

I'm pretty sure it's all installed between his engine and transmission waiting on wiring to go in the car.

[The Dark Side of Will]

Time for a 6 month update

Lol...

This is all I got:

The prototype TOB holder ended up a little small across the throw out finger notches. This allows it to rotate between the throw out fingers more than it should be able to, which messes up snug clearances to Getrag case features. Spec is 2.670.





The original bearing developed some roughness over the course of development, so I had to test the "removal access" holes to use in driving the bearing off so I could replace it with the spare I ordered. R&R were surprisingly easy.



Stock vs dual disk unit



Easiest Install-for-Flight ever!

The TOB is in the transmission and the clutch is on the engine. They'll go together during or after harness build. Harness build is still easier/faster with the engine on the stand than with the cradle built.

[ericjon262]

looks like this thread is more important to me...



the short dip in TPS was the 1-2 Shift, note RPM goes to the limit, and wheel speed doesn't correlate, clutch didn't hold, didn't even slow down...

ERG. I'm feeling out a few options...

[The Dark Side of Will]

Not sure I emphasized it enough previously, but the Northstar build shown is *very lucky* because the stock flexplate has a large enough flat spot to just stack the flywheel on top of it. I have not seen another stock flexplate with this feature.

The F23 has a deeper bellhousing than the 282 and may be able to accommodate a spacer between the flywheel and flexplate. The F40 has a much deeper bellhousing and can definitely accommodate such a spacer. The F40 may even be able to accommodate a dual organic disk clutch.

[Shaun41178(2)]

That garage less guy on YouTube has a twin disk setup from spec that I think is for an f23 in his 3800 swap. Message him